January 13, 2004
14:38
WSPC/101-CEJ
00092
D. M. Fitzgerald, G. A. Zarillo & S. Johnston
586
water and therefore there has been considerable interest in exploring their capabili-
ties for measuring wave spectra. The work of Pinkel and Smith (1987) and Krogstad
et al. (1988) demonstrated that a Doppler sonar using horizontally projected beams
could provide a high quality measurement of wave direction. Whereas, these early
contributions made advancements on various aspects of the problem, they were
not comprehensive, and contained little comparison data to assess the performance
of the ADCP against commonly used wave direction sensors, such as heave-pitch-roll
buoys and pressure velocity (PUV) type sensors.
Recently, Strong et al. (2000) reported on the advancements in ADCP technology
for directional wave measurements and presented comparisons with more conven-
tional wave gauges. In ADCP wave measurements the range cells along the acoustic
beams constitutes a sparse array of independent sensors that measures the local in-
stantaneous velocity projected along the beam. The auto- and cross-spectra of these
signals in each frequency band are assumed to be known linear functionals of the
directional distribution of the waves. Wave direction can be estimated by inverting
this "forward" relation using the Iterative Maximum Likelihood Method (Krogstad
et al., 1988; Terray et al., 1990). The elements of the cross-spectral matrix at any
frequency (or wavenumber) contain directional information in both their phase and
amplitude. Thus, the ADCP lies partway between a pure array measurement, relying
solely on phase, and a point sensor such as a PUV gauge. Figure 16 compares sur-
face wave spectra derived from an InterOcean S4tm sensor with an RDItm 1200 kHz
ADCP unit. The S4 acquires directional wave information using the standard PUV
configuration collecting pressure and velocity data at sampling rate of 2 Hz (Rrbk
and Andersen, 1997). It can be seen that the surface elevation results from the S4
and ADCP unit are nearly identical.
Figure 17 compares the directional energy distribution calculated from the S4
and ADCP sensors. The comparison shows that the Maximum Entropy Method
(MEM) of analysis used to extract directional data from the S4 measurements yields
a broader and smoother directional distribution compared with the Maximum Like-
lihood Method (MLM) analysis of the ADCP measurements. Other comparisons
yielded similar results and indicate that the MEM consistently resolves directional
seas somewhat better than the MLM for both unimodal as well as bimodal seas
(Nwogu et al., 1987). However, Rrbk and Andersen (1997) noted that both meth-
ods require that the wave field is spatially homogeneous, hence effects from diffrac-
tion, refraction, and reflection may influence on the results (due to phase locking).
Under these conditions, Nwogu, et al. (1987) suggest that both methods become
inaccurate with the MLM being the least accurate.
2.8. Geographical information system methods
Aerial photography in digital format is commonly analyzed to study the evolution of
coastal systems. Standard aerial photography is generally of high geometric fidelity
Previous Page
